Control method, device, storage medium and program product for numerically controlled machine tools

Abstract

The present application discloses a control method, device, storage medium and program product for a numerically controlled machine tool, and relates to the field of data processing technology. The present application discloses a control method for a numerically controlled machine tool, which is applied to a first device and includes: obtaining machine tool operation data of a target machine tool; sending the machine tool operation data to a second device, so that the second device inputs the machine tool operation data into a fault detection model to obtain a fault detection result, and when the fault detection result indicates that the target machine tool has a fault, inputs the machine tool operation data into a fault location model to obtain a fault location result and returns it; receives the fault location result, and when the fault location result indicates that a known fault has occurred in the target machine tool, adjusts the operation parameters of the target machine tool based on the fault location result. The present application can improve the efficiency of fault diagnosis for numerically controlled machine tools.

Description

Technical Field

[0001] The present application relates to the field of data processing technology, and in particular to a control method, device, storage medium and program product for a numerically controlled machine tool. Background Art

[0002] CNC machine tools are automated machine tools that use digital computer technology to control the movement and processing of machine tools. Compared with traditional machine tools, CNC machine tools have higher processing accuracy, flexibility and automation, and can process a variety of complex parts. They are widely used in many manufacturing fields such as automobiles, aerospace, and precision instruments. If a CNC machine tool fails, it will directly cause the machine tool to stop working, affecting the continuity and production efficiency of the production line and causing production delays.

[0003] At present, the methods for diagnosing CNC machine tool faults can usually be intuitive inspection and rule matching. The intuitive inspection method is often through the maintenance personnel's observation of the machine tool, and judging the operating status and fault conditions of the machine tool based on their own experience. This method relies on the experience and subjective factors of the maintenance personnel and is inefficient in fault diagnosis. The rule matching method is to analyze the operating rules and fault characteristics of the machine tool, manually record the conditions of the fault, and establish a corresponding rule base. When a fault occurs, the machine tool fault is diagnosed by matching the rule base based on the real-time collected machine tool status information. This method requires the fault to be retrieved from the rule base containing a large amount of data, resulting in the efficiency of fault diagnosis being affected. In summary, the current efficiency of fault diagnosis for CNC machine tools is low.

[0004] The above contents are only used to assist in understanding the technical solution of the present application and do not constitute an admission that the above contents are prior art. Summary of the invention

[0005] The main purpose of the present application is to provide a control method, device, storage medium and program product for a CNC machine tool, aiming to solve the technical problem of low efficiency in fault diagnosis of CNC machine tools.

[0006] To achieve the above object, the present application proposes a control method for a CNC machine tool, which is applied to a first device, wherein the first device is communicatively connected to a second device, and a fault detection model and a fault location model are deployed in the second device, and the method includes:

[0007] Obtaining machine tool operation data of the target machine tool;

[0008] The machine tool operation data is sent to the second device, so that the second device inputs the machine tool operation data into the fault detection model to obtain a fault detection result. When the fault detection result is that the target machine tool has a fault, the machine tool operation data is input into the fault location model to obtain a fault location result and return it, wherein the fault detection model is trained with historical operation data as input data and historical detection results as training labels, and the fault location model is trained with historical operation data as input data and historical location results as training labels;

[0009] The fault location result is received, and when the fault location result indicates that a known fault occurs in the target machine tool, an operating parameter of the target machine tool is adjusted based on the fault location result.

[0010] In one embodiment, the step of adjusting the operating parameters of the target machine tool based on the fault location result includes:

[0011] Determining a fault parameter among the operating parameters of the target machine tool based on the fault location result;

[0012] Determining the preset target parameter data corresponding to the fault parameter from the preset theoretical parameter data corresponding to different operating parameters;

[0013] The fault parameter is adjusted according to the target parameter data.

[0014] In one embodiment, after the step of receiving the fault location result, the method further includes:

[0015] When the fault location result indicates that an unknown fault occurs in the target machine tool, obtaining a fault repair measure of a maintenance personnel, and determining a fault location and a fault cause based on the fault repair measure;

[0016] The fault location and the fault cause are sent to the second device, so that the second device can update the fault location result through the fault location and the fault cause to obtain an updated fault location result, use the machine tool operation data as input data, use the updated fault location result as a training label to update the fault location model, and use the updated fault location model as the fault location model.

[0017] In addition, to achieve the above-mentioned purpose, the present application proposes a control method for a CNC machine tool, which is applied to a second device, in which a fault detection model and a fault location model are deployed, and the second device is communicatively connected with the first device, and the method includes:

[0018] Receiving machine tool operation data of a target machine tool, wherein the machine tool operation data is acquired and sent by the first device;

[0019] Inputting the machine tool operation data into the fault detection model to obtain a fault detection result, wherein the fault detection model is trained using historical operation data as input data and historical detection results as training labels;

[0020] In the case where the fault detection result is that the target machine tool has a fault, the machine tool operation data is input into the fault location model to obtain a fault location result, wherein the fault location model is trained using the historical operation data as input data and the historical location results as training labels;

[0021] The fault location result is sent to the first device, so that the first device can adjust the operating parameters of the target machine tool based on the fault location result when the fault location result indicates that a known fault occurs in the target machine tool.

[0022] In one embodiment, the fault detection model includes a first input layer, a first hidden layer and a first output layer, the first hidden layer includes a function layer, and the function layer includes a plurality of nonlinear activation functions with variable parameters;

[0023] The step of inputting the machine tool operation data into the fault detection model to obtain a fault detection result comprises:

[0024] Sending the machine tool operation data to the first input layer to obtain first input data;

[0025] Sending the first input data to the first hidden layer, splitting the first input data into a plurality of node data in the first hidden layer according to the number of nonlinear activation functions in the function layer, inputting the plurality of node data into each nonlinear activation function of the function layer to obtain an output data sequence, and superimposing each data in the output data sequence to obtain hidden layer output data;

[0026] The hidden layer output data is input into the first output layer to obtain a fault detection result.

[0027] In one embodiment, the machine tool operation data includes operation parameter data of a plurality of operation parameters; the first hidden layer includes a plurality of cascaded function layers, each level of the function layer includes a plurality of function units, and the number of nonlinear activation functions in each function unit of the same level is the same;

[0028] The step of splitting the first input data into a plurality of node data in the first hidden layer according to the number of nonlinear activation functions in the function layer, inputting the plurality of node data into each nonlinear activation function of the function layer to obtain an output data sequence, and superimposing each data in the output data sequence to obtain hidden layer output data comprises:

[0029] For any target parameter data in each of the operating parameter data, the target input data is split according to the target number in the first hidden layer to obtain a plurality of node data corresponding to the target input data, wherein the target input data is the first input data corresponding to the target parameter data, and the target number is the number of nonlinear activation functions in the function unit of the first-level function layer;

[0030] Inputting a plurality of node data corresponding to the target input data into each nonlinear activation function of the target function unit to obtain a target data sequence, wherein the target function unit is any function unit in the first hierarchical function layer;

[0031] The first data sequence is obtained by respectively superimposing the data with the same sequence number in the target data sequence and the non-target data sequence, wherein the non-target data sequence is obtained by respectively inputting a plurality of node data corresponding to the non-target input data into each non-linear activation function of the non-target function unit, the plurality of node data corresponding to the non-target input data is obtained by splitting the non-target input data according to the target number, the non-target input data is the first input data corresponding to the non-target parameter data, the non-target parameter data is the data in each of the operating parameter data except the target parameter data, and the non-target function unit is the function unit in the first hierarchical function layer except the target function unit;

[0032] Each data in the first data sequence is used as input data of the next level function layer until the last level function layer is reached, and the output data of each function unit in the last level function layer is superimposed to obtain the hidden layer output data.

[0033] In one embodiment, the fault location model includes a second input layer, a feature extraction layer, a second hidden layer, and a second output layer, wherein the second hidden layer includes a machine tool fault knowledge base and an inference engine;

[0034] The step of inputting the machine tool operation data into the fault location model to obtain a fault location result comprises:

[0035] Sending the machine tool operation data to the second input layer to obtain second input data;

[0036] Sending the second input data to the feature extraction layer to obtain second feature data;

[0037] Inputting the second feature data into the second hidden layer, querying the machine tool fault knowledge base in the second hidden layer to obtain a fault case, and sending the fault case to the inference engine to obtain fault information;

[0038] The fault information is sent to the second output layer to obtain a fault location result.

[0039] In addition, to achieve the above-mentioned purpose, the present application also proposes a control device for a CNC machine tool, the device comprising: a memory, a processor, and a computer program stored in the memory and executable on the processor, the computer program being configured to implement the steps of the control method for the CNC machine tool as described above.

[0040] In addition, to achieve the above-mentioned purpose, the present application also proposes a storage medium, which is a computer-readable storage medium, and a computer program is stored on the storage medium. When the computer program is executed by the processor, the steps of the control method of the CNC machine tool as described above are implemented.

[0041] In addition, to achieve the above-mentioned purpose, the present application also provides a computer program product, which includes a computer program, and when the computer program is executed by a processor, the steps of the control method of the CNC machine tool as described above are implemented.

[0042] In the present application, a first device obtains machine tool operation data of a target machine tool; sends the machine tool operation data to the second device so that the second device inputs the machine tool operation data into the fault detection model to obtain a fault detection result; when the fault detection result indicates that a fault has occurred in the target machine tool, the machine tool operation data is input into the fault location model to obtain a fault location result and returns it, wherein the fault detection model is trained using historical operation data as input data and historical detection results as training labels, and the fault location model is trained using historical operation data as input data and historical location results as training labels; receives the fault location result, and when the fault location result indicates that a known fault has occurred in the target machine tool, adjusts the operation parameters of the target machine tool based on the fault location result.

[0043] In this application, the real-time operation data of the target machine tool is collected by automated means without the need for regular manual inspection, thereby improving the efficiency of data acquisition and thus improving the overall efficiency of subsequent fault diagnosis.

[0044] In the fault detection process, the fault detection model is first used to determine whether there is a fault based on the machine tool operation data, and then the fault location model is used to accurately locate the machine tool fault information based on the machine tool operation data. Compared with relying on manual experience identification for fault diagnosis, this application reduces the manual operation process and improves the efficiency of fault diagnosis. At the same time, compared with using rule matching for fault diagnosis, this application does not need to compare the machine tool operation data with the preset rules one by one, but uses a machine learning algorithm to automatically learn from the operation data and extract fault-related features for fault diagnosis, reducing the operation steps and operation time in the fault diagnosis process, thereby improving the efficiency of fault diagnosis.

[0045] In addition, in the present application, the fault information is directly located through the fault location model. If the fault occurring in the machine tool is a known fault type, the machine tool operating parameters are automatically adjusted according to the fault location result, and the operating data is processed and analyzed. Then, feature extraction and pattern recognition and other technologies are used to determine whether the machine tool has a fault. The present application can directly lead to machine tool maintenance, avoiding the process of manual repeated debugging of the machine tool for maintenance after feature extraction and pattern recognition, thereby improving the overall efficiency of fault diagnosis and recovery.

[0046] In addition, compared to deploying the fault detection model and the fault location model on the first device, the present application deploys the complex fault detection model and the fault location model on the second device, separates the model calculation process and production control, avoids the impact of the model calculation speed and real-time performance due to the production control process occupying computing resources, and improves the efficiency and accuracy of fault diagnosis for CNC machine tools. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the present application and, together with the description, serve to explain the principles of the present application.

[0048] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, for ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative labor.

[0049] Figure 1 A flow chart of the first embodiment of the control method of the numerical control machine tool of the present application;

[0050] Figure 2 A flow chart of the second embodiment of the control method for a numerically controlled machine tool of the present application;

[0051] Figure 3 A schematic diagram of the structure of the hidden layer in the fault detection model provided in the second embodiment of the present application;

[0052] Figure 4 A schematic diagram of a control process of a numerically controlled machine tool provided in one embodiment of the present application;

[0053] Figure 5 It is a schematic diagram of the equipment structure of the hardware operating environment involved in the control method of the CNC machine tool in the embodiment of the present application.

[0054] The purpose, features and advantages of this application will be further described in conjunction with the embodiments and with reference to the accompanying drawings. DETAILED DESCRIPTION

[0055] It should be understood that the specific embodiments described herein are only used to explain the technical solutions of the present application and are not used to limit the present application.

[0056] In order to better understand the technical solution of the present application, a detailed description will be given below in conjunction with the accompanying drawings and specific implementation methods.

[0057] The main solution of the embodiment of the present application is: the first device obtains the machine tool operation data of the target machine tool; the machine tool operation data is sent to the second device, so that the second device inputs the machine tool operation data into the fault detection model to obtain a fault detection result, and when the fault detection result is that the target machine tool has a fault, the machine tool operation data is input into the fault location model to obtain a fault location result and returned; the fault location result is received, and when the fault location result indicates that a known fault has occurred in the target machine tool, the operating parameters of the target machine tool are adjusted based on the fault location result.

[0058] In this embodiment, for the convenience of description, the following description is made by taking the control device of a numerically controlled machine tool as the execution subject.

[0059] Traditional CNC machine tool fault diagnosis methods usually include the following:

[0060] The visual inspection method is often to observe the machine tool through maintenance personnel, and judge the operating status and fault conditions of the machine tool based on their own experience. This method relies on the experience and subjective factors of the maintenance personnel, and is inefficient in fault diagnosis. The rule matching method is to analyze the operating rules and fault characteristics of the machine tool, manually record the conditions of the fault, and establish a corresponding rule base. When a fault occurs, the machine tool fault is diagnosed by matching the rule base based on the real-time collected machine tool status information. This method requires the retrieval of the fault from the rule base containing a large amount of data, which affects the efficiency of fault diagnosis.

[0061] Signal analysis method: This method processes and analyzes the signal information collected during the operation of the machine tool, and diagnoses the machine tool fault through feature extraction and pattern recognition. Although this method can automatically identify the fault type, it requires a lot of time and effort to locate the cause of the fault and perform fault repairs, which affects the efficiency of fault diagnosis and machine tool recovery.

[0062] In the embodiment of the present application, the real-time operating data of the target machine tool is collected by automated means without the need for regular manual inspection, thereby improving the efficiency of data acquisition and thus improving the overall efficiency of subsequent fault diagnosis.

[0063] In the fault detection process, the fault detection model is first used to determine whether there is a fault based on the machine tool operation data, and then the fault location model is used to accurately locate the machine tool fault information based on the machine tool operation data. Compared with relying on manual experience to identify faults for diagnosis, the embodiment of the present application reduces the manual operation process and improves the efficiency of fault diagnosis. At the same time, compared with using rule matching for fault diagnosis, the embodiment of the present application does not need to compare the machine tool operation data with the preset rules one by one, but uses a machine learning algorithm to automatically learn from the operation data and extract fault-related features for fault diagnosis, which reduces the operation steps and operation time in the fault diagnosis process, thereby improving the efficiency of fault diagnosis.

[0064] Furthermore, in the embodiment of the present application, the fault information is directly located through the fault location model. If the fault occurring in the machine tool is a known fault type, the machine tool operating parameters are automatically adjusted according to the fault location result, and the operating data is processed and analyzed. Then, feature extraction and pattern recognition and other technologies are used to determine whether the machine tool has a fault. The embodiment of the present application can directly lead to machine tool maintenance, avoiding the process of repeated manual debugging of the machine tool for maintenance after feature extraction and pattern recognition, thereby improving the overall efficiency of fault diagnosis and recovery.

[0065] Compared with deploying the fault detection model and the fault location model on the first device, the present application deploys the complex fault detection model and the fault location model on the second device, separates the model calculation process from the production control, avoids the model calculation speed and real-time performance being affected by the production control process occupying computing resources, and improves the efficiency and accuracy of fault diagnosis for CNC machine tools. In addition, it can avoid occupying the computing resources of the first device, reduce the operating burden of the first device, and allow the first device to focus on the core production control function, thereby improving the efficiency of the first device in controlling CNC machine tools.

[0066] It should be noted that the execution subject of this embodiment can be a computing service device with data processing, network communication and program running functions, such as a tablet computer, a personal computer, a mobile phone, etc., or an electronic device capable of realizing the above functions, a control device for a CNC machine tool, etc. The following takes the first device as an example to illustrate this embodiment and the following embodiments.

[0067] Based on this, the embodiment of the present application provides a control method for a CNC machine tool, which is applied to a first device, wherein the first device is communicated with a second device respectively, and a fault detection model and a fault location model are deployed in the second device. Figure 1 This is a flow chart of the first embodiment of the control method of a CNC machine tool of the present application.

[0068] In this embodiment, the control method of the CNC machine tool includes steps S10 to S30.

[0069] Step S10, obtaining machine tool operation data of the target machine tool.

[0070] It should be noted that the target machine tool can be any CNC machine tool, and the machine tool operation data is the data collected by the target machine tool during its operation. Specifically, it can be the real-time data of the machine tool during its operation collected by various sensors (such as vibration sensors, temperature sensors, current sensors, etc.) installed on the CNC machine tool. The machine tool operation data includes but is not limited to the vibration amplitude, temperature change, motor current intensity, etc. of each component of the machine tool. That is, the machine tool operation data is the data that can reflect the real-time working status of the machine tool.

[0071] The first device is communicatively connected with various sensors installed on the CNC machine tool to receive machine tool operation data of the target machine tool from the sensors. In a specific implementation, the first device can collect machine tool operation data from the sensors according to a preset period, or can collect machine tool operation data after receiving instructions in fault diagnosis. The specific settings can be made according to actual needs and are not limited here.

[0072] The first device collects various key data of the CNC machine tool during operation in real time or regularly, including but not limited to temperature, vibration, current, etc., monitors the machine tool status in real time, and provides a basis for subsequent fault analysis.

[0073] In one feasible implementation, after collecting the sensor data, the first device can perform preprocessing such as data cleaning, normalization and feature extraction, and use the preprocessed data as machine tool operation data. The preprocessing can remove noise, standardize the data format, and extract features that are meaningful for fault detection, thereby improving the accuracy and efficiency of data analysis.

[0074] Step S20, sending the machine tool operation data to the second device, so that the second device inputs the machine tool operation data into the fault detection model to obtain a fault detection result. When the fault detection result is that the target machine tool has a fault, the machine tool operation data is input into the fault location model to obtain a fault location result and return it, wherein the fault detection model is trained using historical operation data as input data and historical detection results as training labels, and the fault location model is trained using historical operation data as input data and historical location results as training labels.

[0075] In this embodiment, the first device sends the machine tool operation data to the second device, and a fault detection model and a fault location model are deployed on the second device, wherein the fault detection model uses the historical operation data as input data and the historical detection results corresponding to the historical operation data as training labels for training, and the fault location model uses the historical operation data as input data and the historical location results as training labels for training. The structures of the fault detection model and the fault location model are not limited here and can be set according to actual needs. It should be noted that the historical operation data can be the operation data of the target machine tool or the operation data of the same type of machine tool as the target machine tool. The historical detection results represent whether the machine tool corresponding to the historical operation data has a fault, and the historical location results represent the fault information of the machine tool corresponding to the historical operation data, including but not limited to the faulty component, the cause of the fault, etc.

[0076] The fault detection model and the fault location model can be trained on the second device, or deployed on the second device after being trained on a device other than the first device and the second device. Compared with model training on the second device, deploying the model on the second device after being trained on a device other than the first device and the second device can reduce the operating load of the second device and improve the operating efficiency of the model in the second device, thereby improving the efficiency of fault diagnosis for CNC machine tools.

[0077] After receiving the machine tool operation data, the second device inputs the machine tool operation data into the pre-trained fault detection model to obtain a fault detection result, which is used to characterize whether the target machine tool has a fault. The fault detection model can identify abnormal patterns in the data by learning patterns in historical data, so that after receiving the machine tool operation data, it can determine whether the target machine tool has a fault based on the learned knowledge. In this embodiment, the use of the model for automated fault detection reduces the reliance on manual inspection and improves the detection speed and accuracy of fault detection for CNC machine tools.

[0078] When it is determined that the fault detection model determines that the machine tool has a fault, the second device inputs the machine tool operation data into the fault location model to obtain a fault location result, which represents the fault information of the target machine tool. The fault detection model uses historical data and its corresponding fault location results as training labels, and can identify the source or specific location of the fault from the current operation data, so that when the machine tool operation data is received, the specific fault information of the target machine tool can be determined based on the learned knowledge and the machine tool operation data.

[0079] The second device sends the fault detection result to the first device, and the first device adjusts the CNC machine tool according to the fault detection result.

[0080] Step S30 , receiving the fault location result, and adjusting the operating parameters of the target machine tool based on the fault location result when the fault location result indicates that a known fault occurs in the target machine tool.

[0081] The first device receives the fault location result and, when the fault of the CNC machine tool is a known fault, automatically adjusts the operating parameters of the machine tool according to the fault type of the CNC machine tool, including but not limited to changing the feed speed, spindle speed, coolant flow, etc., to reduce the impact of the fault, avoid further damage, or directly take measures to restore the machine tool to normal operation. The first device can quickly restore the normal working state of the machine tool through timely and effective parameter adjustment, reduce production interruptions caused by faults, and improve overall production efficiency and equipment utilization.

[0082] In this embodiment, the real-time operation data of the target machine tool is collected by automated means without the need for manual periodic inspection, thereby improving the efficiency of data acquisition and thus improving the overall efficiency of subsequent fault diagnosis.

[0083] In the fault detection process, the fault detection model is first used to determine whether there is a fault based on the machine tool operation data, and then the fault location model is used to accurately locate the machine tool fault information based on the machine tool operation data. Compared with relying on manual experience identification for fault diagnosis, this embodiment reduces the manual operation process and improves the efficiency of fault diagnosis. At the same time, compared with using rule matching for fault diagnosis, this embodiment does not need to compare the machine tool operation data with the preset rules one by one, but uses a machine learning algorithm to automatically learn from the operation data and extract fault-related features for fault diagnosis, reducing the operation steps and operation time in the fault diagnosis process, thereby improving the efficiency of fault diagnosis.

[0084] Furthermore, in this embodiment, the fault information is directly located through the fault location model. If the fault occurring in the machine tool is a known fault type, the machine tool operating parameters are automatically adjusted according to the fault location result, and the operating data is processed and analyzed. Then, feature extraction and pattern recognition and other technologies are used to determine whether the machine tool has a fault. This embodiment can directly lead to machine tool maintenance, avoiding the process of repeated manual debugging of the machine tool for maintenance after feature extraction and pattern recognition, thereby improving the overall efficiency of fault diagnosis and recovery.

[0085] Compared with deploying the fault detection model and the fault location model on the first device, the present application deploys the complex fault detection model and the fault location model on the second device, separates the model calculation process from the production control, avoids the model calculation speed and real-time performance being affected by the production control process occupying computing resources, and improves the efficiency and accuracy of fault diagnosis for CNC machine tools. In addition, it can avoid occupying the computing resources of the first device, reduce the operating burden of the first device, and allow the first device to focus on the core production control function, thereby improving the efficiency of the first device in controlling CNC machine tools.

[0086] In one implementation, step S30: adjusting the operating parameters of the target machine tool based on the fault location result may include steps S301 to S303.

[0087] Step S301 : determining fault parameters in the operating parameters of the target machine tool based on the fault location result.

[0088] It should be noted that the fault parameters refer to the parameters of the target machine tool's operating parameters that deviate from the normal range.

[0089] In this embodiment, the first device uses the fault location result to identify the specific fault cause that causes the abnormal operation of the target machine tool. Specifically, it can be through comparing and analyzing the various operating parameters of the machine tool (such as rotation speed, feed rate, cutting force, etc.) and the deviation between the actual performance and the expected performance. Once the root cause of the fault is determined, it can be determined which operating parameters deviate from the normal range, thereby obtaining the fault parameters.

[0090] Step S302, determining the preset target parameter data corresponding to the fault parameter from the preset theoretical parameter data corresponding to different operating parameters.

[0091] It should be noted that the theoretical parameter data refers to the parameter values ​​of various operating parameters when the target machine tool is in normal operating state. Different operating parameters have different values ​​and are not limited here. The theoretical parameter data can be obtained based on the equipment manufacturer's recommendations, historical optimal operating conditions or industry standards. Theoretical parameter data can guide how to adjust parameters to achieve optimal performance or solve specific faults.

[0092] After identifying the fault parameters, the first device determines the preset target parameter data corresponding to the fault parameters from the preset theoretical parameter data corresponding to different operating parameters, so as to ensure that the subsequent parameter adjustment process is based on scientific basis and practical experience, which helps to restore the machine tool to the design performance or a better state and reduce the cost of trial and error.

[0093] Step S303: adjusting the fault parameters according to the target parameter data.

[0094] After obtaining the target parameter data, the first device executes parameter modification through the automated control system. During the adjustment process, the target parameter value must be accurately matched, and the machine tool operating status is monitored at the same time to ensure that the problem can be effectively solved after the adjustment and no new faults will be introduced.

[0095] In this embodiment, by directly adjusting the parameters of the CNC machine tool when it is determined that a known fault occurs in the CNC machine tool, the efficiency and accuracy of fault diagnosis and processing are improved, and the unplanned downtime of the CNC machine tool is reduced. In addition, by directly adjusting the faulty CNC machine tool, the reliability of the production system is enhanced, the production delays caused by the fault are reduced, and the overall production efficiency of the CNC machine tool is improved.

[0096] In one implementation, after the step S10: receiving the fault location result, the method further includes steps S40-S50.

[0097] Step S40, when the fault location result indicates that an unknown fault occurs in the target machine tool, obtaining the fault repair measures of the maintenance personnel, and determining the fault location and the fault cause based on the fault repair measures.

[0098] When the fault location model cannot identify a specific known fault type, that is, when it is determined to be an unknown fault, the maintenance personnel will be notified to perform maintenance, and the maintenance personnel's fault repair measures will be obtained. It should be noted that the fault repair measures can be obtained by monitoring the maintenance personnel's operations through sensors, or by receiving fault repair measures input by maintenance personnel based on the visualization page of the control system. Fault repair measures can also be obtained in other feasible ways, which are not limited here and can be set according to actual needs.

[0099] The first device receives the fault repair measures input by the maintenance personnel, and determines the fault location and fault cause based on the fault repair measures. The specific method of determining the fault location and fault cause is not limited here. For example, the fault location and fault cause can be determined by keyword extraction. Based on the fault repair measures, the approximate location and possible causes of the fault can be determined, providing direction for subsequent maintenance and model optimization, so that when the fault occurs again in the future, measures can be taken to initially locate the fault, laying the foundation for fault resolution, while also accumulating new cases for the knowledge base to promote continuous learning and improvement.

[0100] Step S50, sending the fault location and the fault cause to the second device, so that the second device updates the fault location result through the fault location and the fault cause to obtain an updated fault location result, using the machine tool operation data as input data and the updated fault location result as a training label to update the fault location model, and using the updated fault location model as the fault location model.

[0101] After determining the fault location and the fault cause, the first device sends the fault location and the fault cause to the second device.

[0102] The second device updates the fault location result through the fault part and the fault cause to obtain an updated fault location result, and then uses the machine tool operation data as input data and the updated fault location result as a training label to update the fault location model, and uses the updated fault location model as the fault location model. By incorporating cases of unknown faults into the learning scope of the model, the model can be retrained to more accurately identify and locate such faults when encountering similar situations in the future.

[0103] In this implementation, a dynamic model update mechanism is used to enable the fault location model to continuously evolve, thereby improving its ability to identify unknown or new faults, reducing the misjudgment rate of similar faults in the future, enhancing the system's self-learning and adaptability, improving the accuracy and efficiency of fault diagnosis, reducing production losses caused by faults, optimizing maintenance resource allocation, and improving the overall operation and maintenance level.

[0104] Based on the first embodiment of the present application, in the second embodiment of the present application, the same or similar contents as those in the above-mentioned embodiment 1 can be referred to the above introduction, and no further description will be given later. In this embodiment, the control method of the CNC machine tool is applied to the second device, the fault detection model and the fault location model are deployed in the second device, and the second device is connected to the first device in communication. On this basis, please refer to Figure 2 , the method described includes steps A10~A30.

[0105] Step A10: receiving machine tool operation data of the target machine tool, wherein the machine tool operation data is acquired and sent by the first device.

[0106] In this embodiment, the first device acquires the machine tool operation data of the target machine tool and sends the data to the second device, and the second device receives the machine tool operation data of the target machine tool.

[0107] It should be noted that the target machine tool can be any CNC machine tool, and the machine tool operation data is the data collected by the target machine tool during its operation. Specifically, it can be the real-time data of the machine tool during its operation collected by various sensors (such as vibration sensors, temperature sensors, current sensors, etc.) installed on the CNC machine tool. The machine tool operation data includes but is not limited to the vibration amplitude, temperature change, motor current intensity, etc. of each component of the machine tool. That is, the machine tool operation data is the data that can reflect the real-time working status of the machine tool.

[0108] Step A20, inputting the machine tool operation data into the fault detection model to obtain a fault detection result, wherein the fault detection model is trained using historical operation data as input data and historical detection results as training labels.

[0109] The second device inputs the machine tool operation data into the fault detection model to obtain a fault detection result, which is used to characterize whether the target machine tool has a fault. The fault detection model can identify abnormal patterns in the data by learning patterns in historical data, so that after receiving the machine tool operation data, it can determine whether the target machine tool has a fault based on the learned knowledge. In this embodiment, the use of the model for automated fault detection reduces the reliance on manual inspection and improves the detection speed and accuracy of fault detection for CNC machine tools.

[0110] Step A30, when the fault detection result is that the target machine tool has a fault, the machine tool operation data is input into the fault location model to obtain a fault location result, wherein the fault location model is trained using the historical operation data as input data and the historical location results as training labels.

[0111] When it is determined that the fault detection model determines that the machine tool has a fault, the second device inputs the machine tool operation data into the fault location model to obtain a fault location result, which represents the fault information of the target machine tool. The fault detection model uses historical data and its corresponding fault location results as training labels, and can identify the source or specific location of the fault from the current operation data, so that when the machine tool operation data is received, the specific fault information of the target machine tool can be determined based on the learned knowledge and the machine tool operation data.

[0112] Step A40: Send the fault location result to the first device, so that the first device can adjust the operating parameters of the target machine tool based on the fault location result when the fault location result indicates that a known fault occurs in the target machine tool.

[0113] The second device sends the fault detection result to the first device, and the first device adjusts the CNC machine tool according to the fault detection result. Specifically, the first device receives the fault location result, and when the fault of the CNC machine tool is a known fault, it will automatically adjust the operating parameters of the machine tool according to the fault type of the CNC machine tool, including but not limited to changing the feed speed, spindle speed, coolant flow, etc., to reduce the impact of the fault, avoid further damage, or directly take measures to restore the machine tool to normal operation. The first device can quickly restore the normal working state of the machine tool through timely and effective parameter adjustment, reduce production interruptions caused by faults, and improve overall production efficiency and equipment utilization.

[0114] It should be noted that the specific implementation of steps A10 to A40 in this embodiment can refer to steps S10 to S30, which will not be described in detail here.

[0115] In this embodiment, the real-time operation data of the target machine tool is collected by automated means without the need for manual periodic inspection, thereby improving the efficiency of data acquisition and thus improving the overall efficiency of subsequent fault diagnosis.

[0116] In the fault detection process, the fault detection model is first used to determine whether there is a fault based on the machine tool operation data, and then the fault location model is used to accurately locate the machine tool fault information based on the machine tool operation data. Compared with relying on manual experience identification for fault diagnosis, this embodiment reduces the manual operation process and improves the efficiency of fault diagnosis. At the same time, compared with using rule matching for fault diagnosis, this embodiment does not need to compare the machine tool operation data with the preset rules one by one, but uses a machine learning algorithm to automatically learn from the operation data and extract fault-related features for fault diagnosis, reducing the operation steps and operation time in the fault diagnosis process, thereby improving the efficiency of fault diagnosis.

[0117] Furthermore, in this embodiment, the fault information is directly located through the fault location model. If the fault occurring in the machine tool is a known fault type, the machine tool operating parameters are automatically adjusted according to the fault location result, and the operating data is processed and analyzed. Then, feature extraction and pattern recognition and other technologies are used to determine whether the machine tool has a fault. This embodiment can directly lead to machine tool maintenance, avoiding the process of repeated manual debugging of the machine tool for maintenance after feature extraction and pattern recognition, thereby improving the overall efficiency of fault diagnosis and recovery.

[0118] Compared with deploying the fault detection model and the fault location model on the first device, the present application deploys the complex fault detection model and the fault location model on the second device, separates the model calculation process from the production control, avoids the model calculation speed and real-time performance being affected by the production control process occupying computing resources, and improves the efficiency and accuracy of fault diagnosis for CNC machine tools. In addition, it can avoid occupying the computing resources of the first device, reduce the operating burden of the first device, and allow the first device to focus on the core production control function, thereby improving the efficiency of the first device in controlling CNC machine tools.

[0119] In one embodiment, the fault detection model includes a first input layer, a first hidden layer and a first output layer, the first hidden layer includes a function layer, and the function layer includes a plurality of nonlinear activation functions with variable parameters. Step A20, inputting the machine tool operation data into the fault detection model to obtain a fault detection result, includes steps A201 to A203.

[0120] Step A201, sending the machine tool operation data to the first input layer to obtain the first input data.

[0121] The neural network architecture used in the fault detection model includes an output layer (hereinafter referred to as the first input layer for distinction), a hidden layer (hereinafter referred to as the first hidden layer for distinction) and an output layer (hereinafter referred to as the first output layer for distinction). The first hidden layer uses a nonlinear activation function with variable parameters, which is the key to the neural network's ability to learn complex patterns. The nonlinear activation function can introduce nonlinear relationships, allowing the model to fit a wider range of data distributions, thereby better capturing the nonlinear association between faults and operating data.

[0122] The machine tool operation data is first passed to the first input layer, and after preprocessing (such as standardization and normalization), it is converted into a form that can be processed by the model to form the first input data. The input layer allows the original data to adapt to the input requirements of the model, ensuring the consistency of the data format with the model requirements. It is the basis for subsequent analysis and improves the effectiveness of the model in processing data.

[0123] Step A202: Send the first input data to the first hidden layer, split the first input data into multiple node data in the first hidden layer according to the number of nonlinear activation functions in the function layer, input the multiple node data into each nonlinear activation function of the function layer to obtain an output data sequence, and superimpose each data in the output data sequence to obtain hidden layer output data.

[0124] After the first input data enters the first hidden layer, the first hidden layer splits the first input data into multiple node data according to the number of nonlinear activation functions in the function layer, and each node data is processed by a nonlinear activation function with variable parameters to generate a new feature representation, and the new feature identifiers of each node data are combined into a sequence, that is, the output data sequence.

[0125] In one embodiment, the machine tool operation data includes operation parameter data of multiple operation parameters; the first hidden layer includes multiple cascaded function layers, each level of the function layer includes multiple function units, and the number of nonlinear activation functions in each function unit of the same level is the same. In this embodiment, in the first hidden layer, the first input data is split into multiple node data according to the number of nonlinear activation functions in the function layer, and the multiple node data are respectively input into each nonlinear activation function of the function layer to obtain an output data sequence. The process of superimposing each data in the output data sequence to obtain the hidden layer output data can be:

[0126] a1. For any target parameter data in each of the operating parameter data, the target input data is split according to the target quantity in the first hidden layer to obtain multiple node data corresponding to the target input data, wherein the target input data is the first input data corresponding to the target parameter data, and the target quantity is the number of nonlinear activation functions in the function unit of the first-level function layer.

[0127] Machine tool operation data may cover multiple important parameters monitored by the CNC machine tool during operation, such as temperature, vibration frequency, motor current, etc. Each item is a type of operation parameter data, and they together constitute a comprehensive description of the machine tool status.

[0128] For any operating parameter data (referred to as target parameter data), the second device splits the target input data in the first hidden layer according to the target quantity to obtain multiple node data corresponding to the target input data. It should be noted that the target input data is the first input data corresponding to the target parameter data, and the target quantity is the number of nonlinear activation functions in the function unit of the first-level function layer.

[0129] a2. Input the multiple node data corresponding to the target input data into each nonlinear activation function of the target function unit to obtain a target data sequence, wherein the target function unit is any function unit in the first-level function layer.

[0130] The multiple node data corresponding to the target input data are respectively input into each nonlinear activation function of the target function unit to obtain an output data sequence, which is hereinafter referred to as the target data sequence for distinction. It should be noted that the target function unit is any function unit in the first-level function layer. In this embodiment, the use of nonlinear mathematical transformation enhances the model's ability to learn complex data patterns and improves the model's accuracy in identifying complex fault patterns.

[0131] a3, respectively superimposing the data with the same sequence number in the target data sequence and the non-target data sequence to obtain a first data sequence, wherein the non-target data sequence is obtained by respectively inputting a plurality of node data corresponding to the non-target input data into each non-linear activation function of the non-target function unit, the plurality of node data corresponding to the non-target input data is obtained by splitting the non-target input data according to the target quantity, the non-target input data is the first input data corresponding to the non-target parameter data, the non-target parameter data is the data in each of the operating parameter data except the target parameter data, and the non-target function unit is the function unit in the first-level function layer except the target function unit.

[0132] It should be noted that the data other than the target parameter data in each operating parameter data is called non-target input data, the multiple node data corresponding to the non-target input data are obtained by splitting the non-target input data according to the target number, and the output data sequence obtained by inputting the multiple node data corresponding to each non-target input data into each non-linear activation function of the non-target function unit is called a non-target data sequence, and the non-target function unit is a function unit other than the target function unit in the first-level function layer.

[0133] The target data sequence and the data with the same sequence number in the non-target data sequence are superimposed respectively to obtain a first data sequence. The method of superimposing the target data sequence and the non-target data sequence is not limited herein, and can be a weighted summation or a direct addition. In a specific implementation, direct addition can be used to reduce the amount of model calculation.

[0134] In this embodiment, please refer to Figure 3 The schematic diagram of the structure of the hidden layer in the fault detection model shown, Figure 3In the example, the operating parameter data includes two types of operating parameter data. The first hidden layer includes two function layers. The first level function layer includes two function units, each function unit includes 5 nonlinear activation functions, and the second level function layer includes five function units, each function unit includes 1 nonlinear activation function. In the first hidden layer, the nonlinear activation function is used as a model parameter that can be learned. Therefore, the parameters of the nonlinear activation function used in this embodiment are variable, such as Spline (spline function) and sigmoid function (logic function). In the training process of the fault detection model, the nonlinear activation function is updated as a parameter that needs to be learned. Figure 3 In the hidden layer structure shown, when the output data of the previous function layer is used as the input data of the next function layer, until the preset number of layers is reached, the output data of each function unit is superimposed to obtain the hidden layer output data.

[0135] Traditional multi-layer perceptrons are usually composed of linear transformations (i.e., weight matrices multiplied by input vectors) and fixed nonlinear activation functions. Although adding hidden layers can improve the expressive power of the model, the fixed activation function limits the network's ability to explore complex function spaces. However, in this embodiment, the nonlinear activation function is used as a learnable parameter in the model, allowing the model to adaptively learn the most suitable nonlinear transformation method for the current task, so that the model not only learns the weight matrix, but also learns how to optimally transform the input data, so that it can more flexibly approximate any complex functional relationship, thereby improving the accuracy of the model, and then improving the accuracy of the fault detection results.

[0136] Step A203, inputting the hidden layer output data into the first output layer to obtain a fault detection result.

[0137] The hidden layer output data is sent to the first output layer, which is responsible for making the final fault judgment based on the extracted features. The output layer may use a classification algorithm to obtain a prediction result of fault or no fault, or it may use a logistic regression algorithm to obtain the failure probability of the target machine tool, which is not restricted here.

[0138] In this embodiment, the fault detection model realizes in-depth analysis of CNC machine tool operation data and accurate fault identification through layered processing of neural networks, especially the application of nonlinear activation functions. This process not only improves the efficiency and accuracy of fault detection, but also enhances the adaptability to unknown or complex faults through learning and continuous optimization, which has significant beneficial effects on maintaining the stable operation of CNC machine tools, reducing downtime, and improving production efficiency.

[0139] In one embodiment, the fault localization model includes a second input layer, a feature extraction layer, a second hidden layer and a second output layer, and the second hidden layer includes a machine tool fault knowledge base and an inference engine; step A30, inputting the machine tool operation data into the fault localization model to obtain a fault localization result, including steps A301~A304.

[0140] Step A301, sending the machine tool operation data to the second input layer to obtain second input data.

[0141] The fault location model combines deep learning with knowledge-driven methods and consists of a second input layer, a feature extraction layer, a second hidden layer, and a second output layer. This design aims to process data features through deep learning and use the knowledge base for fault logic reasoning to achieve accurate fault location.

[0142] The machine tool operation data is first sent to the second input layer, which is responsible for performing necessary preprocessing on the original data, such as standardization and normalization, to ensure that the data is suitable for subsequent feature extraction, ensure the quality of the model input data, and improve the model learning efficiency and accuracy.

[0143] Step A302, sending the second input data to the feature extraction layer to obtain second feature data.

[0144] The preprocessed data enters the feature extraction layer, which extracts features that are meaningful for fault location from the data through various feature extraction techniques (such as convolution, pooling, self-attention mechanism, etc.) to form the second feature data. The extracted features can better reflect the fault mode in the data and provide key information for subsequent fault identification.

[0145] Step A303, input the second feature data into the second hidden layer, query the machine tool fault knowledge base in the second hidden layer to obtain fault cases, and send the fault cases to the inference engine to obtain fault information.

[0146] The second feature data is fed into the second hidden layer, which contains the machine tool fault knowledge base and the reasoning engine. The knowledge base stores a large number of known fault cases and solutions. The reasoning engine queries the knowledge base based on the input feature data, matches the corresponding fault cases, and then uses the case information for logical reasoning to obtain specific fault information.

[0147] Step A304, sending the fault information to the second output layer to obtain the fault location result.

[0148] The fault information is transmitted to the second output layer, which is responsible for organizing the fault information and outputting it as the final fault location result, clearly indicating the location and possible causes of the fault, and providing a clear and direct fault diagnosis report, so that maintenance personnel can take quick measures, shorten the fault repair time, and reduce downtime losses.

[0149] In this embodiment, the fault location model realizes efficient conversion from data to fault information by integrating the automatic feature extraction capability of deep learning and knowledge-based reasoning analysis, thereby improving the accuracy of fault location. If the fault of the machine tool is a known fault type, the machine tool operating parameters are automatically adjusted according to the fault location result, and the operating data is processed and analyzed. Then, feature extraction and pattern recognition and other technologies are used to determine whether the machine tool has a fault. This embodiment can directly lead to machine tool maintenance, avoiding the process of manual repeated debugging of the machine tool for maintenance after feature extraction and pattern recognition, thereby improving the overall efficiency of fault diagnosis and recovery.

[0150] Exemplarily, in order to help understand the technical concept or technical principle of the control method of the CNC machine tool after the present embodiment is combined with the above-mentioned embodiment 1 and embodiment 2, in the present embodiment, a control method of the CNC machine tool based on multi-agent collaboration is proposed, and multi-agent (intelligent body) is introduced to collaborate with each other in the fault diagnosis of the CNC machine tool, which significantly reduces the time for manual analysis of machine tool abnormalities, positioning abnormal causes, and finding solutions, thereby effectively reducing the downtime of the machine tool during the production process and reducing the impact on production. In the present embodiment, the control process of the CNC machine tool involves 5 agents: a data acquisition agent, a data processing agent, a fault detection agent, a fault diagnosis agent, and a fault response agent. Each agent can be deployed on the same device or on different devices. For example, the data acquisition agent, the data processing agent, and the fault response agent can be deployed on the first device, and the fault detection agent and the fault diagnosis agent can be deployed on the second device. It should be noted that a fault detection model is deployed in the fault detection agent, and a fault location model is deployed on the fault diagnosis agent. In the present embodiment, refer to Figure 4 , the specific control process of CNC machine tools can be:

[0151] 1. First, the data acquisition agent continuously monitors the operating status of the machine tool through a network of sensors connected to the machine tool. These sensors can capture key performance indicators such as temperature, vibration, and current in real time and transmit this data to the data processing agent (i.e. Figure 4 The first device shown collects the machine tool operation data of the target machine tool through the data collection Agent, that is, obtains the machine tool operation data of the target machine tool).

[0152] 2. After the data processing agent receives the data, it starts to perform preprocessing tasks such as data cleaning, normalization, and feature extraction to remove noise, standardize the data format, and extract features that are meaningful for fault detection, thereby improving the accuracy and efficiency of data analysis.

[0153] 3. When the preprocessed data is transmitted to the fault detection agent, the agent will use the Kolmogorov–Arnold Networks (KAN) algorithm (also known as the fault detection model) to analyze whether the data deviates from the normal mode, that is, to detect whether anomalies have occurred. Once an anomaly is found, the relevant data will be sent to the fault diagnosis agent for further analysis (also known as Figure 4 The second device shown inputs the machine tool operation data into the fault detection model to obtain a fault detection result).

[0154] 4. When the abnormal data is transmitted to the fault diagnosis agent, the agent will use the machine tool fault knowledge base based on the Large Language Model (LLM) to analyze and interpret the detected abnormal data. By comparing the known fault type with the current data, the fault type and possible cause can be quickly determined, and corresponding maintenance suggestions can be provided. This information is then passed to the fault response agent (i.e. Figure 4 As shown, when it is determined that the fault detection result is that the target machine tool has a fault, the second device inputs the machine tool operation data into the fault location model to obtain a fault location result).

[0155] 5. The fault response agent will send a notification to the operator or administrator according to the information transmitted by the fault diagnosis agent, or directly adjust the machine tool parameters to alleviate or solve the problem, ensuring that the machine tool can quickly resume production (i.e. Figure 4 As shown, in the case where the fault location result indicates that a known fault occurs in the target machine tool, the first device adjusts the operating parameters of the target machine tool based on the fault location result).

[0156] 6. When encountering an unknown fault type, the model self-updating agent will be activated. This agent will automatically collect relevant data when the fault occurs and the measures taken to solve the fault, and use this new information to update the machine tool fault knowledge base, so that the system can more quickly identify and solve the same or similar fault problems in future diagnosis (i.e. Figure 4 As shown, when the fault location result indicates that an unknown fault has occurred in the target machine tool, the first device obtains fault repair measures and determines the fault location and the cause of the fault based on the fault repair measures. The first device sends the fault location and the cause of the fault to the second device. The second device updates the fault location result according to the fault location and the cause of the fault to obtain an updated fault location result, uses the machine tool operation data as input data, uses the updated fault location result as a training label to update the fault location model, and uses the updated fault location model as the fault location model).

[0157] It should be noted that the above examples are only used to understand the present application and do not constitute a limitation on the control method of the CNC machine tool of the present application. More simple transformations based on this technical concept are all within the protection scope of the present application.

[0158] The present application provides a control device for a CNC machine tool. The control device for the CNC machine tool can be a first device or a second device. The control device for the CNC machine tool includes: at least one processor; and a memory connected to the at least one processor for communication; wherein the memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute the control method for the CNC machine tool in the above-mentioned embodiment one.

[0159] Reference below Figure 5 , which shows a schematic diagram of the structure of a control device for a numerically controlled machine tool suitable for implementing an embodiment of the present application. The control device for a numerically controlled machine tool in the embodiment of the present application may include, but is not limited to, mobile terminals such as mobile phones, laptop computers, digital broadcast receivers, PDAs (Personal Digital Assistants), PADs (Portable Application Descriptions), PMPs (Portable Media Players), vehicle-mounted terminals (such as vehicle-mounted navigation terminals), etc., and fixed terminals such as digital TVs, desktop computers, etc. Figure 5 The control device of the CNC machine tool shown is only an example and should not bring any limitation to the functions and scope of use of the embodiments of the present application.

[0160] like Figure 5As shown, the control device of the numerical control machine tool may include a processing device 1001 (such as a central processing unit, a graphics processor, etc.), which can perform various appropriate actions and processes according to the program stored in the read-only memory (ROM: Read Only Memory) 1002 or the program loaded from the storage device 1003 to the random access memory (RAM: Random Access Memory) 1004. In the RAM 1004, various programs and data required for the operation of the control device of the numerical control machine tool are also stored. The processing device 1001, the ROM 1002 and the RAM 1004 are connected to each other through a bus 1005. An input / output (I / O) interface 1006 is also connected to the bus. Generally, the following systems can be connected to the I / O interface 1006: an input device 1007 including, for example, a touch screen, a touch pad, a keyboard, a mouse, an image sensor, a microphone, an accelerometer, a gyroscope, etc.; an output device 1008 including, for example, a liquid crystal display (LCD: Liquid Crystal Display), a speaker, a vibrator, etc.; a storage device 1003 including, for example, a magnetic tape, a hard disk, etc.; and a communication device 1009. The communication device 1009 can allow the control device of the numerical control machine tool to communicate with other devices wirelessly or by wire to exchange data. Although the control device of the numerical control machine tool with various systems is shown in the figure, it should be understood that it is not required to implement or have all the systems shown. More or fewer systems can be implemented or provided instead.

[0161] In particular, according to the embodiments disclosed in the present application, the process described above with reference to the flowchart can be implemented as a computer software program. For example, the embodiments disclosed in the present application include a computer program product, which includes a computer program carried on a computer-readable medium, and the computer program includes a program code for executing the method shown in the flowchart. In such an embodiment, the computer program can be downloaded and installed from a network through a communication device, or installed from a storage device 1003, or installed from a ROM 1002. When the computer program is executed by the processing device 1001, the above-mentioned functions defined in the method of the embodiment disclosed in the present application are executed.

[0162] The control device of the numerically controlled machine tool provided by the present application adopts the control method of the numerically controlled machine tool in the above-mentioned embodiment, and can solve the technical problem of low efficiency in fault diagnosis of the numerically controlled machine tool. Compared with the prior art, the beneficial effects of the control device of the numerically controlled machine tool provided by the present application are the same as the beneficial effects of the control method of the numerically controlled machine tool provided by the above-mentioned embodiment, and other technical features in the control device of the numerically controlled machine tool are the same as the features disclosed in the method of the previous embodiment, and will not be repeated here.

[0163] It should be understood that the various parts disclosed in this application can be implemented by hardware, software, firmware or a combination thereof. In the description of the above embodiments, specific features, structures, materials or characteristics can be combined in any one or more embodiments or examples in a suitable manner.

[0164] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art who is familiar with the present technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

[0165] The present application provides a computer-readable storage medium having computer-readable program instructions (ie, computer programs) stored thereon, and the computer-readable program instructions are used to execute the control method of the numerically controlled machine tool in the above-mentioned embodiment.

[0166] The computer-readable storage medium provided in the present application may be, for example, a USB flash drive, but is not limited to electrical, magnetic, optical, electromagnetic, infrared, or semiconductor systems, systems or devices, or any combination of the above. More specific examples of computer-readable storage media may include, but are not limited to: an electrical connection with one or more wires, a portable computer disk, a hard disk, a random access memory (RAM: Random Access Memory), a read-only memory (ROM: Read Only Memory), an erasable programmable read-only memory (EPROM: Erasable Programmable Read Only Memory or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM: CD-Read Only Memory), an optical storage device, a magnetic storage device, or any suitable combination of the above. In this embodiment, the computer-readable storage medium may be any tangible medium containing or storing a program, which may be used by or in combination with an instruction execution system, system or device. The program code contained on the computer-readable storage medium may be transmitted using any appropriate medium, including but not limited to: wires, optical cables, RF (Radio Frequency: Radio Frequency), etc., or any suitable combination of the above.

[0167] The computer-readable storage medium may be included in the control device of the CNC machine tool; or may exist independently without being assembled into the control device of the CNC machine tool.

[0168] The computer-readable storage medium carries one or more programs. When the one or more programs are executed by the control device of the CNC machine tool, the control device of the CNC machine tool executes the control method of the CNC machine tool in the above embodiment.

[0169] Computer program code for performing the operations of the present application may be written in one or more programming languages ​​or a combination thereof, including object-oriented programming languages ​​such as Java, Smalltalk, C++, and conventional procedural programming languages ​​such as "C" or similar programming languages. The program code may be executed entirely on the user's computer, partially on the user's computer, as a separate software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In the case of a remote computer, the remote computer may be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or may be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0170] The flow chart and block diagram in the accompanying drawings illustrate the possible architecture, function and operation of the system, method and computer program product according to various embodiments of the present application. In this regard, each square box in the flow chart or block diagram can represent a module, a program segment or a part of a code, and the module, the program segment or a part of the code contains one or more executable instructions for realizing the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the square box can also occur in a sequence different from that marked in the accompanying drawings. For example, two square boxes represented in succession can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each square box in the block diagram and / or flow chart, and the combination of the square boxes in the block diagram and / or flow chart can be implemented with a dedicated hardware-based system that performs a specified function or operation, or can be implemented with a combination of dedicated hardware and computer instructions.

[0171] The modules involved in the embodiments described in this application may be implemented by software or hardware, wherein the name of the module does not constitute a limitation on the unit itself in some cases.

[0172] The readable storage medium provided in the present application is a computer-readable storage medium, which stores computer-readable program instructions (i.e., computer programs) for executing the control method of the above-mentioned CNC machine tool, and can solve the technical problem of low efficiency in fault diagnosis of CNC machine tools. Compared with the prior art, the beneficial effects of the computer-readable storage medium provided in the present application are the same as the beneficial effects of the control method of the CNC machine tool provided in the above-mentioned embodiment, and will not be described in detail here.

[0173] The present application also provides a computer program product, comprising a computer program, wherein when the computer program is executed by a processor, the steps of the control method of the CNC machine tool as described above are implemented.

[0174] The computer program product provided by the present application can solve the technical problem of low efficiency in fault diagnosis of CNC machine tools. Compared with the prior art, the beneficial effects of the computer program product provided by the present application are the same as the beneficial effects of the control method of the CNC machine tool provided by the above embodiment, which will not be described in detail here.

[0175] The above descriptions are only some embodiments of the present application, and are not intended to limit the patent scope of the present application. All equivalent structural changes made using the contents of the present application specification and drawings under the technical concept of the present application, or direct / indirect applications in other related technical fields are included in the patent protection scope of the present application.

Claims

1. A control method for a CNC machine tool, It is characterized in that Applied to a first device, the first device is communicatively connected to a second device respectively, a fault detection model and a fault location model are deployed in the second device, and the method includes: Obtaining machine tool operation data of the target machine tool; The machine tool operation data is sent to the second device, wherein the second device inputs the machine tool operation data into the fault detection model to obtain a fault detection result. When the fault detection result is that the target machine tool has a fault, the machine tool operation data is input into the fault location model to obtain a fault location result and return it. The fault detection model is trained with historical operation data as input data and historical detection results as training labels. The fault location model is trained with historical operation data as input data and historical location results as training labels. The fault location result is used to characterize the fault information of the target machine tool, and the fault information includes the fault location and the fault cause. receiving the fault location result, and adjusting the operating parameters of the target machine tool based on the fault location result when the fault location result indicates that a known fault occurs in the target machine tool; Wherein, after the step of receiving the fault location result, the method further includes: When the fault location result indicates that an unknown fault occurs in the target machine tool, obtaining a fault repair measure of a maintenance personnel, and determining a fault location and a fault cause based on the fault repair measure; The fault location and the fault cause are sent to the second device, so that the second device updates the fault location result according to the fault location and the fault cause to obtain an updated fault location result, uses the machine tool operation data as input data, uses the updated fault location result as a training label to update the fault location model, and uses the updated fault location model as the fault location model; The machine tool operation data includes operation parameter data of a plurality of operation parameters; the fault detection model includes a first input layer, a first hidden layer and a first output layer, the first hidden layer includes a plurality of cascaded function layers, each level of the function layer includes a plurality of function units, and the number of nonlinear activation functions in each function unit of the same level is the same; The step of inputting the machine tool operation data into the fault detection model by the second device to obtain a fault detection result comprises: The second device sends the machine tool operation data to the first input layer to obtain first input data; For any target parameter data in each of the operating parameter data, the target input data is split according to the target number in the first hidden layer to obtain a plurality of node data corresponding to the target input data, wherein the target input data is the first input data corresponding to the target parameter data, and the target number is the number of nonlinear activation functions in the function unit of the first-level function layer; Inputting a plurality of node data corresponding to the target input data into each nonlinear activation function of the target function unit to obtain a target data sequence, wherein the target function unit is any function unit in the first hierarchical function layer; The first data sequence is obtained by respectively superimposing the data with the same sequence number in the target data sequence and the non-target data sequence, wherein the non-target data sequence is obtained by respectively inputting a plurality of node data corresponding to the non-target input data into each non-linear activation function of the non-target function unit, the plurality of node data corresponding to the non-target input data is obtained by splitting the non-target input data according to the target number, the non-target input data is the first input data corresponding to the non-target parameter data, the non-target parameter data is the data in each of the operating parameter data except the target parameter data, and the non-target function unit is the function unit in the first hierarchical function layer except the target function unit; Using each data in the first data sequence as input data of the next level function layer until reaching the last level function layer, and superimposing the output data of each function unit in the last level function layer to obtain hidden layer output data; The hidden layer output data is input into the first output layer to obtain a fault detection result.

2. The method according to claim 1, It is characterized in that The step of adjusting the operating parameters of the target machine tool based on the fault location result includes: Determining a fault parameter among the operating parameters of the target machine tool based on the fault location result; Determining the preset target parameter data corresponding to the fault parameter from the preset theoretical parameter data corresponding to different operating parameters; The fault parameter is adjusted according to the target parameter data.

3. A control method for a CNC machine tool, It is characterized in that Applied to a second device, a fault detection model and a fault location model are deployed in the second device, and the second device is communicatively connected with the first device, and the method includes: Receiving machine tool operation data of a target machine tool, wherein the machine tool operation data is acquired and sent by the first device; Inputting the machine tool operation data into the fault detection model to obtain a fault detection result, wherein the fault detection model is trained using historical operation data as input data and historical detection results as training labels; In the case where the fault detection result is that the target machine tool has a fault, the machine tool operation data is input into the fault location model to obtain a fault location result, wherein the fault location model is trained using the historical operation data as input data and the historical location results as training labels, and the fault location result is used to characterize the fault information of the target machine tool, and the fault information includes the fault location and the fault cause; sending the fault location result to the first device, wherein the first device adjusts the operating parameters of the target machine tool based on the fault location result when the fault location result indicates that a known fault occurs in the target machine tool; The machine tool operation data includes operation parameter data of a plurality of operation parameters; the fault detection model includes a first input layer, a first hidden layer and a first output layer, the first hidden layer includes a plurality of cascaded function layers, each level of the function layer includes a plurality of function units, and the number of nonlinear activation functions in each function unit of the same level is the same; The step of inputting the machine tool operation data into the fault detection model to obtain a fault detection result comprises: Sending the machine tool operation data to the first input layer to obtain first input data; For any target parameter data in each of the operating parameter data, the target input data is split according to the target number in the first hidden layer to obtain a plurality of node data corresponding to the target input data, wherein the target input data is the first input data corresponding to the target parameter data, and the target number is the number of nonlinear activation functions in the function unit of the first-level function layer; Inputting a plurality of node data corresponding to the target input data into each nonlinear activation function of the target function unit to obtain a target data sequence, wherein the target function unit is any function unit in the first hierarchical function layer; The first data sequence is obtained by respectively superimposing the data with the same sequence number in the target data sequence and the non-target data sequence, wherein the non-target data sequence is obtained by respectively inputting a plurality of node data corresponding to the non-target input data into each non-linear activation function of the non-target function unit, the plurality of node data corresponding to the non-target input data is obtained by splitting the non-target input data according to the target number, the non-target input data is the first input data corresponding to the non-target parameter data, the non-target parameter data is the data in each of the operating parameter data except the target parameter data, and the non-target function unit is the function unit in the first hierarchical function layer except the target function unit; Using each data in the first data sequence as input data of the next level function layer until reaching the last level function layer, and superimposing the output data of each function unit in the last level function layer to obtain hidden layer output data; Inputting the hidden layer output data into the first output layer to obtain a fault detection result; Wherein, the method further comprises: The first device obtains a maintenance personnel's fault repair measure when the fault location result indicates that the target machine tool has an unknown fault, and determines the fault location and fault cause based on the fault repair measure; The fault location and the fault cause are sent to the second device, so that the second device can update the fault location result through the fault location and the fault cause to obtain an updated fault location result, use the machine tool operation data as input data, use the updated fault location result as a training label to update the fault location model, and use the updated fault location model as the fault location model.

4. The method according to claim 3, It is characterized in that The fault location model includes a second input layer, a feature extraction layer, a second hidden layer and a second output layer, wherein the second hidden layer includes a machine tool fault knowledge base and an inference engine; The step of inputting the machine tool operation data into the fault location model to obtain a fault location result comprises: Sending the machine tool operation data to the second input layer to obtain second input data; Sending the second input data to the feature extraction layer to obtain second feature data; Inputting the second feature data into the second hidden layer, querying the machine tool fault knowledge base in the second hidden layer to obtain a fault case, and sending the fault case to the inference engine to obtain fault information; The fault information is sent to the second output layer to obtain a fault location result.

5. A control device for a CNC machine tool, It is characterized in that The device comprises: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the computer program is configured to implement the steps of the control method for a numerically controlled machine tool according to any one of claims 1 to 4.

6. A storage medium, It is characterized in that The storage medium is a computer-readable storage medium, and a computer program is stored on the storage medium. When the computer program is executed by a processor, the steps of the control method of a numerically controlled machine tool as claimed in any one of claims 1 to 4 are implemented.

7. A computer program product, It is characterized in that The computer program product comprises a computer program, and when the computer program is executed by a processor, the steps of the method for controlling a numerically controlled machine tool according to any one of claims 1 to 4 are implemented.

Images